通过分解提示结构，让大模型自动发现并优化关键推理组件。
Adaptive Prompt Structure Factorization: A Framework for Self-Discovering and Optimizing Compositional Prompt Programs
- 用架构模型挖掘任务专属的提示语义因子，拆解复杂提示。
- 单因子干预更新，准确评估每个部分贡献，提升优化效率。
- 在多个推理任务上提升准确率，节省45%~87%调优成本。
自动化提示优化对激发大语言模型（LLMs）可靠推理至关重要，但现有仅通过API操作的优化器通常迭代修改整体提示，耦合各组件，模糊责任归属，限制可控性并浪费计算资源。我们提出API-only框架aPSF（自适应提示结构分解），利用架构模型发现任务特定的提示语义因子，并进行干预式单因子更新：通过验证性能变化估计各因子的边际贡献，结合错误引导选择当前主要失败源进行优化，实现更高效调优。在多个高级推理基准测试中，aPSF优于强基线，包括原理感知优化器，在平均精度上提升最多达+2.16个百分点，且在MultiArith任务中将优化成本降低45%–87%，仅用1步即达到最优验证性能。
原文摘要 · Abstract (English)
Automated prompt optimization is crucial for eliciting reliable reasoning from large language models (LLMs), yet most API-only prompt optimizers iteratively edit monolithic prompts, coupling components and obscuring credit assignment, limiting controllability, and wasting tokens. We propose Adaptive Prompt Structure Factorization (aPSF), an API-only framework (prompt-in/text-out; no access to model internals) that uses an Architect model to discover task-specific prompt structures as semantic factors. aPSF then performs interventional, single-factor updates: interventional factor-level scoring estimates each factor's marginal contribution via validation-performance changes, and error-guided factor selection routes updates to the current dominant failure source for more sample-efficient optimization. Across multiple advanced reasoning benchmarks, aPSF outperforms strong baselines including principle-aware optimizers, improving accuracy by up to +2.16 percentage points on average, and reduces optimization cost by 45--87% tokens on MultiArith while reaching peak validation in 1 step.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。